Towards a Unified Model for Generating Answers and Explanations in Visual Question Answering

Chenxi Whitehouse, Tillman Weyde, Pranava Madhyastha


Abstract
The field of visual question answering (VQA) has recently seen a surge in research focused on providing explanations for predicted answers. However, current systems mostly rely on separate models to predict answers and generate explanations, leading to less grounded and frequently inconsistent results. To address this, we propose a multitask learning approach towards a Unified Model for Answer and Explanation generation (UMAE). Our approach involves the addition of artificial prompt tokens to training data and fine-tuning a multimodal encoder-decoder model on a variety of VQA-related tasks. In our experiments, UMAE models surpass the prior state-of-the-art answer accuracy on A-OKVQA by 10 15%, show competitive results on OK-VQA, achieve new state-of-the-art explanation scores on A-OKVQA and VCR, and demonstrate promising out-of-domain performance on VQA-X.
Anthology ID:
2023.findings-eacl.126
Volume:
Findings of the Association for Computational Linguistics: EACL 2023
Month:
May
Year:
2023
Address:
Dubrovnik, Croatia
Editors:
Andreas Vlachos, Isabelle Augenstein
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1693–1705
Language:
URL:
https://aclanthology.org/2023.findings-eacl.126
DOI:
10.18653/v1/2023.findings-eacl.126
Bibkey:
Cite (ACL):
Chenxi Whitehouse, Tillman Weyde, and Pranava Madhyastha. 2023. Towards a Unified Model for Generating Answers and Explanations in Visual Question Answering. In Findings of the Association for Computational Linguistics: EACL 2023, pages 1693–1705, Dubrovnik, Croatia. Association for Computational Linguistics.
Cite (Informal):
Towards a Unified Model for Generating Answers and Explanations in Visual Question Answering (Whitehouse et al., Findings 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.findings-eacl.126.pdf
Video:
 https://aclanthology.org/2023.findings-eacl.126.mp4